Building Unified Geospatial Data for Land-Change Modeling—A Case Study in the Area of Richmond, Virginia
Abstract – An effort to build a unified collection of geospatial data for use in land-change modeling (LCM) led to new insights into the requirements and challenges of building an LCM data infrastructure. A case study of data compilation and unification for the Richmond, Va., Metropolitan Statistical Area (MSA) delineated the problems of combining and unifying heterogeneous data from many independent localities such as counties and cities. The study also produced conclusions and recommendations for use by the national LCM community, emphasizing the critical need for simple, practical data standards and conventions for use by localities. This report contributes an uncopyrighted core glossary and a much needed operational definition of data unification.
The U.S. Geological Survey (USGS) Eastern Geographic Science Center (EGSC) required a unified and complete set of spatially referenced land-use and land-cover (LULC) data for use in land-change modeling (LCM) projects for the area including and surrounding the City of Richmond, Va. Because an adequate collection of LULC data was not already available for this area, the EGSC took on the effort to find, acquire, and compile a unified set of geospatial LULC data for Richmond and its surrounding area. The EGSC collected and processed heterogeneous geospatial data for 30 categories of geographic features from more than 20 data producers.
This report presents the data unification effort for the Richmond area as a case study, showing how the general problems of unifying heterogeneous data from independent sources were addressed and either solved or mitigated for the specific case of LULC data in this area. This case study tracks months of effort to acquire and prepare a rich collection of geospatial data for use in process-based land-change modeling, and it presents the work in the form of concrete steps taken to build a unified, consistent, and coherent dataset .
The case study and report generalize the discoveries and conclusions made from working with data for the Richmond area into specific recommendations for the land-change modeling community. These recommendations suggest how the modeling community might use and extend these contributions in order to advance the development of an LCM data infrastructure (National Research Council, 2014, p. 96–98).
The innovations presented in this report include
- a much needed operational definition of data unification,
- the core of a working glossary,
- a starting point for developing data standards and data conventions for localities, and
- an outline of three generalizable methods of geographic information system (GIS) processing.
Donato, D.I., and Shapiro, J.L., 2016, Building unified geospatial data for land-change modeling—A case study in the area of Richmond, Virginia: U.S. Geological Survey Open-File Report 2016‒1176, 84 p., https://doi.org/10.3133/ofr20161176.